Normative Uncertainty Dominates Near-Term Mitigation Policy Decisions in Integrated Assessment Models

Climate change is a wicked problem characterized by high uncertainty, ambiguous goals, and diverse normative preferences among actors, which pose significant challenges to just and effective policy design. Cost-Benefit Integrated Assessment Models (CB-IAMs) are widely used to design global mitigation pathways and play a central role in informing and evaluating climate policies within the IPCC Working Group III. However, CB-IAMs reduce the complexity of climate policymaking into a deterministic, unidimensional policy optimization problem. They typically optimize mitigation policy under a single welfare objective, evaluated from the normative perspective of a single representative agent (typically utilitarianism), and consider only a single reference scenario for optimization. This unidimensional framing obscures normative preferences, leaving policymakers without robust or justice-focused information needed in real-world negotiation contexts.

To address these limitations, we introduce JUSTICE, an IAM framework that integrates methods from decision-making under deep and normative uncertainty and welfare economics. JUSTICE implements a general social welfare function (SWF) that explicates normative preferences across regions, time scales, and climate states. This general SWF enables the optimization of policy pathways consistent with multiple distributive justice principles relevant to climate justice discourse. Using JUSTICE, we reformulate the mitigation policy problem into a multi-objective, multi-justice-framing optimization problem. This optimization setup yields Pareto sets of solutions, one for each distributive justice framing. In addition, we conduct a sensitivity analysis of optimal mitigation levels across three types of uncertainty: stochastic, deep, and normative.

Results show that justice framing strongly shapes both the ambition and distribution of global mitigation pathways. The prioritarian framing recommends deeper emission reductions across all SSPs, placing greater normative weight on the temperature objective than the utilitarian framing. The utilitarian framing distributes the mitigation burden more uniformly across regions, whereas the prioritarian framing allocates greater mitigation responsibility to developed regions. Near-term (2050) mitigation ambition is highly sensitive to normative uncertainty, which explains over 50% of the variance in mitigation levels, followed by deep uncertainties arising from socioeconomic dynamics. Stochastic uncertainty originating from the climate's response to emissions has the least influence on the mitigation actions. We find that justice framing determines the spatial distribution of mitigation, while the selection of policy solutions along the Pareto frontier determines the level of ambition. Normative uncertainty is the dominant factor shaping near-term decisions, whereas deep uncertainty becomes increasingly influential towards the end of the century. 

Overall, our results demonstrate the complex interaction of deep and normative uncertainties in mitigation planning. Explicitly disaggregating conflicting objectives and justice perspectives is essential for understanding the distributional consequences of optimal policies. Our methods also expand the range of possible decisions, clarify trade-offs, and ensure the representation of diverse stakeholder values, thereby directly addressing the tenets of procedural justice. When integrated into CB-IAMs, this approach supports the design of fairer climate policies, increases legitimacy, strengthens stakeholder engagement, and facilitates effective climate negotiations.